Anxieties About Modernization,Concerns About Community,and Fear of Crime

Abstract

In contemporary European criminology, there is a growing understanding of the fear of crime as the consequence of, and a code for, broader social anxieties, the origins of which are usually traced to fundamental social and global transformation processes characteristic of late modernity. Within the large body of papers published on this topic, one can differentiate between two distinct perspectives: a generalized insecurity approach, according to which free-floating, amorphous anxieties about modernization are directly projected onto crime, and an expanded community concern approach, whereby abstract anxieties about social change require the prism of local conditions in order to convert into fear of crime. Here, these two perspectives are examined on the basis of survey data from Austria. The results provide support for both approaches, with slight advantages for the generalized insecurity model. We also demonstrate that large parts of the frequently reported association between concerns about incivility and fear of crime can be traced to their common roots in broader social anxieties. Questions are raised pertaining to the generalizability of the findings and a comparative research agenda is encouraged which acknowledges that pathways into fear of crime may differ from country to country, depending on the sociocultural and political–institutional makeup of a society.

Keywords

fear of crime insecurity comparative criminology

Introduction

Fear of crime is a complex and multifaceted social phenomenon. One recent contribution has tried to grapple with this; Farrall, Jackson, and Gray (2009) outlined an analytical differentiation of two phenotypes of fear of crime, which have to be seen as ideal-typical poles of a continuum. Therein, they delimitate concrete episodes of fear in perceived situations of threat (experiential fear) from a communicative employment of crime as a synonym for a general discontent with the living conditions in late modernity (expressive fear). Expressive fear denotes an overarching negative social sensitivity, which utilizes crime as a symbolically loaded cipher for the discourse on society and its ailments. While the authors tend to locate the determinants of experiential fear in factors close to crime and victimization, the meaning of expressive fear can be grasped only in recourse to a variety of other insecurities whose breeding ground is formed by the social, cultural, economic, and political upheaval processes characterizing contemporary Western societies. Very much in line with this, numerous recent criminological works assume that the versatile anxieties and concerns brought about by rapid and profound social change build the true foundation for crime-related feelings of insecurity (Ewald, 2000; Gadd & Jefferson, 2009; Girling, Loader, & Sparks, 2000; Hirtenlehner, 2006, 2008a, 2009; Hollway & Jefferson, 1997; Jackson, 2004, 2006, 2009; Jefferson & Hollway, 2000; Sessar, 1998, 2003, 2008, 2010; Taylor, Evans, & Fraser, 1996; Taylor & Jamieson, 1998; Walklate & Mythen, 2008). Despite the high level of common ground with regard to viewing unease about social transformation processes as the source of fear of crime, these works differ significantly with regard to their presumptions pertaining to (1) the range of fears that are interwoven with crime and (2) the mechanisms that link other fears or anxieties to fear of crime. Generally, one can differentiate between two “paradigms” (Hough, 2009):

a broader approach, according to which fear of crime is more or less directly rooted in abstract anxieties about modernization—“diffuse anxieties brought on by social and economic changes” (Roberts, Stalans, Indermaur, & Hough, 2003, p. 61)—and which accords largely equivalent significance to global, national, and local fears (generalized insecurity approach), as well as,

a narrower approach, which centers much of its attention on the perception of the social and moral “health” of the community and according to which a deep-seated sense of insecurity fed by wider social change requires the prism of local conditions in order to generate fear of crime (expanded community concern approach).

In the following, both perspectives on the development of expressive fear of crime are presented in greater detail and empirically assessed on the basis of survey data collected in Linz, Austria. Similarities and differences found to exist between the two perspectives will be identified and translated into linear structural equation models (SEMs), which are examined comparatively first and then brought together in a more integrative type of modeling strategy. The article closes with a discussion of the implications of our results for international research on the formation of anxieties about crime.

Two Literatures Linking Social Change and Fear of Crime

We are well aware that there are other approaches of explaining fear of crime. Plenty of studies examine the effects of personal or vicarious victimization, media coverage of crime, neighborhood composition and local crime rates, individual vulnerability, and so forth, on fear of crime (Boers, 2003; Hale, 1996; Warr, 2000). Usually, these studies report unsatisfactory low proportions of explained variance in worries about criminal victimization. Sometimes, when admitting the low explanatory power of their models, the authors of these studies refer to social anxieties as an important but omitted predictor variable (e.g., Hirtenlehner, 2008b; Zarafonitou, 2009). Quantitative research directly assessing the impact of broader anxieties on fear of crime, however, has remained scarce up to now.

What has been written on the connection of social anxieties and crime-related fear is based on the assumption that fears and anxieties generated by social transformation processes translate into fear of crime, which in essence becomes a meta-symbol for social problems and pathologies of life in late modernity. Profound social, economic, and political changes provide the breeding ground upon which crime-related feelings of insecurity can flourish. However, which fears are linked to crime, and through what sorts of mechanisms, is a controversial matter.

A generalized insecurity approach, which is favored especially—but not only—in the German-speaking countries, links directly to the diagnosis of an endemic late-modern sense of insecurity (Ewald, 2000; Hirtenlehner, 2006, 2008a, 2009; Sessar, 1998, 2003, 2008, 2010; Taylor & Jamieson, 1998; Walklate & Mythen, 2008). Following the line of argument developed by Bauman (2006, 2007), Beck (1992), and Giddens (1990, 1991), it is supposed that the perceptible loss of traditional embeddings, securities, and certainties across all areas of life encourages a diffuse sense of uneasiness, in which the various risks and insecurities lose their uniqueness and distinctiveness and blend into a generalized threat. The manifold fears that feed on social, economic, political, and cultural transformations detach themselves from their origins and blur into one contourless whole. This nondescript whole in turn enhances and intensifies everyday worries and fears. In this way, anxiety loses the reference to its actual inceptions. It now seizes upon many things that have nothing in the least to do with its origin and equips them with a connotation which they would not receive otherwise. Only against such a background does it become understandable that people fear for their jobs without being threatened by unemployment personally, that people are afraid of military conflict even when they live in a geopolitically secure area, or that people are afraid of crime although their actual risks of victimization are rather modest.

Carrying this line of thought further, it follows that fear of crime does not manifest itself in social reality as a phenomenon distinguishable from other anxieties. Fear of crime is, on the contrary, inextricably intertwined with broader social and existential anxieties. One can speak of an amalgam of unease where the various fears overlap and permeate one another. With a view to the common root of globally, nationally, and locally shaped worries and fears, fear of crime should be understood as a single facet of a much broader sense of insecurity. Fear of crime surfaces as an expression of an unspecific insecurity nurtured by diffuse existential fears and elusive future-oriented anxieties. These subliminal fears are projected onto crime, which then serves as a tangible embodiment of the risks of modernization which are otherwise difficult to pinpoint. The intrapsychological benefit of this projection can be seen in a “division of fear into portions” (Sessar, 2003; based on Baumann, 1999) or in a “defence against anxiety” (Hollway & Jefferson, 1997; based on Giddens, 1991). By breaking down these abstract anxieties into specific problems, they become nameable, communicable, workable, and sometimes even surmountable.¹ Through the channels of problem displacement, the otherwise unbearable risks of modernization are split down into portions that are individually digestible. Crime thus becomes the smallest common denominator for a series of entirely different—social, cultural, economic, ecological, and political—insecurities: a metaphor for everything that is dissatisfying about the new, altered living conditions in a rapidly changing globalized world.

The previously most direct examination of the tenability of a generalized insecurity model is based on survey data collected in Upper Austria (Hirtenlehner, 2006, 2008a). The statistical analysis was performed as a test of a linear SEM in the form of a second-order CFA, which depicts the hypothesized relationships between the various subdimensions of late-modern anxiety and the subliminal amorphous sense of insecurity in a mathematically correct way. The results show that such a model is highly compatible with the observed data and provides a better fit than a simple disorder model. Results from Vienna which were obtained using a less sophisticated statistical methodology point in the same direction (Hirtenlehner & Karazman-Morawetz, 2004).

The other end of the continuum is marked by approaches that are more focused on the state of the neighborhood and community. The expanded community concern approach, which is primarily advocated in the English-speaking world, emphasizes a mediating role of assessments of local community conditions and perceptions of community decline (Farrall et al., 2009; Garofalo & Laub, 1978; Girling et al., 2000; Jackson, 2004, 2009; Taylor et al., 1996; Tulloch et al., 1998). It posits that diffuse anxieties about modernization influence how people make sense of their immediate environment. Deep-seated concerns about the pace and direction of the omnipresent social and economic transformations make people vulnerable to the perception of disorder and incivility, to concerns about the social and moral cohesion of the community, to worries about community fragmentation, to doubts on the capacity of the community to act collaboratively, and to a skeptical assessment of the cultural stability of the neighborhood in terms of fundamental value priorities—in short, to the perception of a breakdown of trusted local structures. A general sense of unease regarding the consequences of social change promotes a vague, gnawing feeling that something is wrong with the community, a sense that social cohesion and moral consensus are in disarray. Neighborhoods and local environments thus act as barometers on which one can read the state of health of the society. These locally tainted diagnoses, in turn, provide fertile feeding grounds for fear of crime to flourish. A negative assessment of the social and moral state of a community sheds doubt on the predictability of the people one meets in daily life, and on the foreseeability of their behaviors. Observations of unreliability and unpredictability then act to encourage an increased fear of crime. The inversion of the argument implies “that crime often operates as a symbol, expressing (…) a number of other issues, (…), insecurities and anxieties regarding one’s neighborhood, its social makeup and status, its place in the world, and the sense that problems from outside were creeping in” (Jackson, 2004, p. 950).

Empirical investigations of such an expanded community concern approach have been conducted by Farrall et al. (2009) and Jackson (2004). In both cases, they relied on survey data collected in England. Both studies show that anxieties about broader social change are associated with fear of crime; the relationship is however mediated to a large extent by assessments of local community conditions. Wider anxieties about transformation give rise to concerns about community decline, about an erosion of the social cohesion and moral consensus within the neighborhood, which in turn pave the way for the fear of crime to flourish. This perspective is supported by a number of qualitative studies conducted in Great Britain, which were able to verify an almost inextricable interlacing of fear of crime and worries about changes in peoples’ immediate lifeworld (Farrall et al., 2009, p. 123 ff.; Girling et al., 2000; Jefferson & Hollway, 2000; Taylor et al., 1996).

The remainder of this article deals with the question of the empirical validity of these two approaches in urban Austria. For both perspectives, linear SEMs are formulated which are tested comparatively on the basis of survey data from Linz.

Method

Sample

The database for our investigation is a survey of residents of the Upper Austrian city of Linz conducted in April 2011.² According to a quota plan drawing on “age,” “sex,” and “residential area” as combined quota criteria, a total of 750 people over 20 years old received a self-complete questionnaire. The questionnaires were handed to the respondents at their front doors and collected the next day.³ When the target could not be found at home the next day, two more attempts to pick up the inventories were undertaken. Thus, 653 completed questionnaires were obtained, which equals a response rate of 87%. Table 1 shows that in terms of the quota criteria the achieved sample is perfectly representative for the residential population.

Table 1.

Sociodemographic Profile of the Respondents and the Urban Population.

	Respondents		Residential population^a
	Persons	%	Persons	%
Age in years
20–29	121	18	27,444	18
30–39	115	18	26,189	17
40–49	128	20	29,766	19
50–59	100	15	24,706	16
60–69	84	13	20,981	13
70 and older	102	16	26,476	17
Sex
Male	312	48	90,234	48
Female	339	52	99,611	52
Area of residence
Inner city	231	35	66,575	35
Southern districts	266	41	77,309	41
Northern districts	156	24	45,961	24

^a Source: City of Linz: http://www.linz.at/zahlen (Status on January 1, 2011).

Measurement

At the core of our analyses is the pattern of relationships between four dimensions of late-modern insecurity—economic fears, social fears, fear of crime, and concerns about incivility—whose measurement will be outlined in the following section.

Economic Fears

Economic fears describe insecurities concerning the economic and financial risks brought about by processes like globalization, deregulation, or the dismantling of the Fordist welfare state. The items combine concerns about the nation’s economic future with worries about one’s own financial future. The respondents were asked to state how worried they felt concerning each of the risks listed in Table 2 (“There are many different risks and dangers in life. We have compiled some of them. Please indicate to what extent you are worried about each of the following issues!”). The selection of indicators is based on a scale developed by Hirtenlehner (2011) to study the impact of economic fears on attitudes toward punishment. Responses ranged from very worried (5) to not worried at all (1).

Table 2.

Items Used to Measure Economic Fears.

Variable		M	SD	r_i;t _−i
E1	Tax increase	3.37	1.12	.70
E2	Loss of pensions	2.97	1.25	.60
E3	Economic crises	3.18	1.06	.64
E4	Increase in prices	3.50	1.14	.76
E5	Decrease in standard of living	2.88	1.15	.64

Cronbach’s α = .85.

M = arithmetic mean; SD = standard deviation; r_i,t _−i = corrected item-total correlation.

Social Fears

Social fears refer to the unease felt with regard to social and health-related issues. The wording of the introductory question was the same as for economic fears. The individual items were selected to depict concerns about interpersonal relationships and subjective health.⁴ Each of the indicators given in Table 3 was to be assessed on a 5-point rating scale ranging from very worried (5) to not worried at all (1).

Table 3.

Items Used to Measure Social Fears.

Variable		M	SD	r_i;t _−i
S1	Loneliness	2.17	1.18	.62
S2	Serious illness	2.61	1.08	.58
S3	War	2.04	1.03	.48
S4	Loss of friends and acquaintances	2.18	1.12	.61
S5	Deterioration of the health system	3.35	1.13	.44

Cronbach’s α = .77.

M = arithmetic mean; SD = standard deviation; r_i,t _−i = corrected item-total correlation.

Fear of Crime

When attempting to tap into emotional fear of crime, researchers may choose between several conceptual approaches. One may rely on global measures (single-item indicators) or specific measures (multiple-item scales differentiating various offense types), one may focus on the frequency or the intensity of crime-related fear, and one may assess fear in terms of being worried, being concerned, or being afraid, among others (Farrall et al., 2009; Hale, 1996). We employed a well-established multiple-item worry scale. The affective dimension of fear of crime was measured as worry concerning the specific offenses and infringements listed in Table 4. We asked “What can you tell us about personal safety in Linz? Please indicate to what extent you are worried about each of the following incidents!” Responses on a Likert-type answering scale ranged from very worried (5) to not worried at all (1). The rationale behind that measurement strategy is 2-fold: On one hand, compared to frequency measures, drawing on the intensity of fear of crime seems better suited to capture the expressive nature of the concept (Farrall et al., 2009); on the other hand, an SEM approach rests upon multiple-item measurement of theoretical constructs (Kline, 2005).

Table 4.

Items Used to Measure Fear of Crime.

Variable		M	SD	r_i;t _−i
C1	Robbery	2.66	1.14	.77
C2	Assault	2.77	1.13	.78
C3	Theft	3.08	1.10	.78
C4	Abuse and harassment	2.97	1.16	.69
C5	Burglary	2.92	1.16	.68

Cronbach’s α = .89.

M = arithmetic mean; SD = standard deviation; r_i,t _−i = corrected item-total correlation.

Incivilities

The measurement of concerns about “low-level breaches of community standards” (Ferraro, 1995, p. 15) combines physical and social sources of irritation. Physical incivilities refer to a disorderly or neglected physical environment; social incivilities describe disruptive social behaviors. Both components are often found to be closely correlated (LaGrange, Ferraro, & Supancic, 1992; Ross & Jang, 2000; Skogan, 1990; Xu, Fiedler, & Flaming, 2005). Following a common practice in operationalizing perceived incivility (e.g., Jackson, 2004; LaGrange et al., 1992; Rengifo & Bolton, 2012; Sampson & Raudenbush, 1999), we asked respondents to assess how much of a problem the nuisances given in Table 5 were for their area (“There can be various problems in a neighborhood. Please indicate how much of a problem the following things are in your neighborhood!”). The response categories ranged from very big problem (4) to no problem at all (1).

Table 5.

Items Used to Measure Incivilities.

Variable		M	SD	r_i;t _−i
I1	Rubbish or litter on the streets	2.22	.90	.68
I2	Graffiti	2.02	.85	.68
I3	Run-down buildings	1.77	.82	.70
I4	Groups of loitering teenagers	2.07	.93	.69
I5	Drunks on the streets	1.86	.89	.70
I6	Vandalism	2.04	.88	.71
I7	Homeless and beggars	1.64	.86	.64

Cronbach’s α = .89.

M = arithmetic mean; SD = standard deviation; r_i,t _−i = corrected item-total correlation.

Models and Analyses

To test the generalized insecurity perspective, an SEM in the form of a second-order CFA (Byrne, 2001, p. 120 ff.) was specified. Core to the model structure is a diffuse insecurity emblematized by a second-order factor. This amorphous sense of insecurity is understood theoretically as a result of social transitions, but its origins are not further investigated here. The unspecific generalized insecurity is broken down into four first-order factors: economic fears, social fears, fear of crime, and concerns about incivility. This modeling approach is based on the assumption that doubts about the stability of a neighborhood are at least partly to be understood as a materialization of broader anxieties. The conceptualization of concerns about incivility as a manifestation of an abstract sense of insecurity—as is fear of crime—conforms to the notion of Sampson and Raudenbush (1999, p. 608) who see incivility and crime as being merely opposite ends of a “seriousness” continuum. If signs of incivility are indications that administrative laws are being broken or minor offenses are being committed, then they can also serve as a projection area for other anxieties in the same way that serious crimes do.

The arrangement of the elements in a factor structure implies that the various anxieties penetrate one another, but that their intertwining arises from their common roots in a more general unease. No causal order is to be found between the different fears, but they can all be traced back to an insecurity of a higher order, and this is why they appear more or less in bundles in reality.⁵

In the course of the community concern perspective, the four anxieties are arranged to form a conventional structural model (Byrne, 2001, p. 142ff). It is hypothesized that economic and social fears—the two concepts are assumed to be correlated—stimulate unease about unwelcome changes in the local neighborhood which in turn generates a heightened fear of crime. Concerns about community conditions are derived from irritation connected to signs of incivility in the neighborhood. According to Farrall et al. (2009), incivilities operate as a “barometer of social cohesion and moral consensus” (p. 6) and as “lay seismograph of neighbourhood stability and breakdown” (p. 99). They signal the presence of crime, the absence of social control, and not least the erosion of commonly accepted values and norms. As such, they readily offer their services as a proxy measure of concerns about community deterioration and neighborhood decline.

At the heart of the community concern approach is the proposition that the impact of general anxieties about modernization and social change on fear of crime is mediated by the assessment of the local habitat. In line with the widening of the perspective that is expressed by the amendment “expanded,” we also specify direct effects of economic and social anxieties on fear of crime. However, the more these direct effects come to the fore when compared to the indirect connecting line, the closer the community concern model draws to a generalized insecurity perspective. This point, therefore, offers an ideal opportunity to recapitulate the similarities and differences between the two approaches.

At their cores, both models imply that worry about crime must be understood as part of a more general worry about the ills of late-modernity (Gadd & Jefferson 2009, p. 136) and that the root cause of fear of crime can be found in broader anxieties about social change. While the generalized insecurity approach suggests a materialization of free-floating, generalized anxieties through their projection on crime, the community concern approach understands the local habitat as a point of crystallization where the various anxieties condense and become perceptible. Social change is most obviously apparent in the immediate environment, according to this thesis. Unconscious efforts to defend against anxieties threatening the self—a notion founded in depth psychology—are contrasted with cognitively driven perceptions of changes in the local lifeworld as intrapsychic linking mechanism. While a generalized insecurity perspective lends itself to a direct translation of universal system risks of global and national reach, of existential fears and fears of falling into fear of crime; a community concern approach—also in its expanded version—conceives perceived community health as a central filter that mediates the effects of these broader, often hard-to-grasp anxieties on fear of crime. This does not remain without consequences for the assumed mixing ratio of the fears transported by the metaphor “crime.” Although unanimous in their recognition of the expressive nature of the fear of crime, the community concern approach implies that fear of crime is symbolically loaded mainly with assessments of local quality of life and the social and moral makeup of the neighborhood; while a generalized insecurity framework, as a result of its greater permeability to general system risks, remains open for larger proportions of anxieties of national and global reach.

Furthermore, differences arise with respect to assumptions concerning the separability of various dimensions of fear and regarding the significance of causality. The community concern model relies on a unidirectional causal understanding of the relationships between the different anxieties, which implies that the constructs can be discriminated empirically. In contrast, the generalized insecurity model draws on the idea of a reciprocal interpenetration and fusion of the various anxieties. No postulations are made as to one-sided cause-and-effect relationships, instead of this fears of global, national, and local character are understood as inextricably interwoven strands of a rope named insecurity (Smith & Pain, 2009).

The statistical analysis was conducted with Amos 16 (Arbuckle 2007). The maximum likelihood algorithm was used for parameter estimation.

Results

Model Comparison

The goal of data analysis with linear SEMs is to identify a model which, with as little complexity as possible, demonstrates a high compatibility with the observed data. The simplicity, or parsimony, of a model can be determined in terms of the number of estimated parameters or in terms of its degrees of freedom.⁶ The model’s goodness of fit is determined by the discrepancy between the empirical and the reproduced covariance structures.

Beginning with the goodness of fit of the overall models, it can be seen that both models are highly compatible with the data. Both models fit the data very well, according to all applied fit parameters.⁷ In both cases, the goodness-of-fit index (GFI) and the comparative fit index (CFI) exceed the acceptance threshold of .90. The root mean square error of approximation (RMSEA) remains below the upper limit of .08. The ratio of χ² to degrees of freedom is smaller than the critical value of 3 for both models. Sure, the χ² test shows in both cases significant deviations between the empirical and the estimated covariance matrices, but due to its dependency on sample size this does not necessarily indicate an unsatisfactory model fit (Byrne, 2001, p. 81).

The discussion of the substantive findings of model estimation shall be opened with the generalized insecurity model. The relationship between the first-order factors and the second-order factor is of particular interest here. All four subdimensions of insecurity are significantly (p ≤ .001) and very substantially loaded on the diffuse second-order insecurity. Fear of crime is very well reproduced by the higher-order insecurity. The standardized factor loading of .82 corresponds to a proportion of explained variance amounting to 67%.⁸ The comparatively lowest factor loading was estimated for social fears—with .59 it is nevertheless high enough to warrant its conceptualization as manifestation of an amorphous sense of insecurity. Concerns about incivility and economic fears take their places between the poles. With factor loadings around .67, they fit very well in the generalized insecurity model. Measured in terms of the observed factor loading parameters, we may conclude that the pattern of relationships between the concepts derived from the generalized insecurity approach is corroborated empirically.

At this juncture, a short note on the goodness of fit of the generalized insecurity model for different subgroups of respondents is warranted. Table 6 displays fit measures for subsamples split by gender, age, and personal victimization background. It is apparent that the model structure derived from the generalized insecurity perspective fits the data for males and females, younger and older respondents, and people who have or have not fallen victim to a crime in the year before the survey. The group-specific analyses yield almost identical fit parameters, which suggests that the compatibility of the model with the data is not due to an extreme match in a particular subpopulation.

Table 6.

Goodness-of-Fit of the Generalized Insecurity Model for Various Subgroups.

	Persons	p	χ²/df	GFI	CFI	RMSEA
Males	297	.000	1.90	.90	.95	.06
Females	312	.000	2.34	.88	.93	.07
Age < 50 years	348	.000	2.17	.90	.94	.06
Age ≥ 50 years	261	.000	2.24	.88	.92	.07
Victims	312	.000	2.04	.90	.94	.06
Nonvictims	293	.000	2.09	.89	.93	.06

Note. CFI = comparative fit index; GFI = goodness of fit; RMSEA = root mean square error of approximation.

An inspection of the inner workings of the community concern model also provides an indication of a compatibility of the postulated relationships with the data. As expected, economic and social fears condense into a pessimistic evaluation of the social and moral state of the neighborhood, which then creates a climate in which fear of crime can thrive. Some doubts pertaining to the appropriateness of a community concern perspective, however, arise from the strength of the direct effects of economic and social anxieties on fear of crime. Concerns about economic risks (β = .28) and worries about social issues (β = .23) exercise a considerable direct influence on fear of crime, which is only slightly inferior to the immediate effect of concerns about incivility (β = .36). Turning to the total effects⁹ of the three predictors, we can see that all in all economic fears (TE = .41) arouse more fear of crime than concerns about incivility (TE = β = .36) do. The total effect of social fears (TE = .30) falls below the one of irritations connected to incivility. The fact that considerable segments of the enchainment of fear of crime with wider anxieties circumvent the assessment of local neighborhood conditions challenges the tenability of a community concern perspective in this data set.

The journey through the isolated results of the different SEMs culminates in the question of which of the two models is, in the end, superior. Such a decision is not easily made. Mathematically, the two SEMs are largely equivalent. Consequently, the two models deliver nearly identical measures of goodness of fit (χ², GFI, CFI, and RMSEA).

Therefore, trying to determine discrepancies in overall fit by means of a χ² difference test¹⁰ does not help one terribly much. The above-mentioned mathematical closeness of the models results in very similar χ² values, which cannot be expected to vary systematically. As a consequence, the more complex community concern model cannot provide a significant increase in goodness of fit compared to the more parsimonious generalized insecurity model, ▵χ² = 4.30; ▵df = 2; p= .117.

Figure 1.

Theoretical models.

Figure 2.

Empirical structural equation models (standardized coefficients). All estimated parameters are significant with p ≤ .001 (CR_min = 3.88).¹¹

Figure 3.

Integrative structural equation model (standardized coefficients).

If the absolute compatibility of the statistically reproduced covariances with the observed data does not provide much help, then the next logical step is to concentrate on the complexity of the model. In principle, the more parsimonious and therewith simpler model would be considered preferable—in that this model would regularly permit a simpler understanding of the relationships inherent in the data. The generalized insecurity model requires and estimates two parameters less than the community concern model, which provides the former with a slight advantage in terms of parsimony.

A comparison of the goodness of fit of different models is usually performed on the basis of information criteria of model fit. Here, the Akaike Information Criterion (AIC; Akaike, 1987) and the Bayes Information Criterion (BIC; Schwarz, 1978) were applied. In both cases, smaller values indicate a better fit. The AIC supports the expanded community concern model; whereas the BIC suggests that the generalized insecurity model is the more advantageous selection. That the two information criteria favor different models is because the BIC penalizes increasing model complexity more severely than the AIC.¹²

Furthermore, it should be noted that the variance in fear of crime is better reproduced by the generalized insecurity model than by the expanded community concern model. While the generalized insecurity model can explain 67% of the variance in fear of crime, the expanded community concern model is “only” capable of explaining 46% of the variance in this fear (which is still a considerable amount, of course). One can infer from that, therefore, that the conceptualization of fear of crime as an immediate projection of an amorphous general sense of insecurity enables a better understanding of the extent and distribution of fear of crime than an approach that pushes the perception of local quality of life into the foreground as a buffer.

Model Integration

With all the necessary caution in the interpretation of our results, on balance our model comparison seems to slightly favor the generalized insecurity perspective. The existence of linkages between economic and social fears on one hand and fear of crime on the other hand that circumvent perceptions of local quality of life, the reduced complexity, and the greater explanatory power of the generalized insecurity model as well as its compatibility with earlier findings from urban Austria (Hirtenlehner, 2006, 2008a; Hirtenlehner & Karazman-Morawetz, 2004) support the assumption that fear of crime serves as a cipher for the discourse on society and its ailments in general, one which extends far beyond an assessment of the integrity of one’s neighborhood and community. But although expressive fear of crime seems to gain its significance in a global access to contemporary risks and a filtration through local upheaval is not exactly necessary, a conflation of the two perspectives can have its merits.

When considering opportunities for model integration, the exact nature of the association between unease about incivility and fear of crime comes to the fore. To approximate this problem empirically, we expanded the second-order CFA representing the generalized insecurity perspective with an additional direct pathway from concerns about incivility on fear of crime. Specification of such an enlarged model is warranted by the fact that the correlation between perceptions of incivility and fear of crime observed so often in international research (Boers, 1991; Brunton-Smith, 2011; Farrall et al., 2009; Rengifo & Bolton, 2012; Skogan, 1990) is basically consistent both with the generalized insecurity and with the community concern perspective. The integrative model enables us to examine whether a conception of worries about crime and incivility as parallel materializations of a free-floating sense of insecurity suffices to capture the intertwining of these two constructs¹³ or whether we have, beyond that, to take other connecting lines between perceptions of incivility and fear of crime into consideration, too.

Given the results of the estimation of such an integrated insecurity model, it is hard to come to a definitive decision. The integrative model shows an excellent fit to the data. It is true that in terms of a χ² difference test (▵χ² = 3.67; ▵df = 1; p = .055) no significant improvement of model fit can be achieved, but the required significance level is missed only barely. The various information criteria do not lead to a clear decision: The AIC favors the integrative model; the BIC supports the generalized insecurity model. From a substantive point of view, a significant, but modest direct effect of concerns about incivility on fear of crime emerges (β = .16; CR = 2.31).

All in all, the findings suggest that the well-established connection between worries about incivility and crime is to a large extent due to their common coloring by an overarching amorphous sense of insecurity. Profound transformation anxieties which are not easily put in concrete terms are responsible for the almost inextricable interweaving between concerns about incivility and fear of crime. The covariation of incivility- and crime-related worries that is not reproduced by this spurious relationship remains comparatively low. According to the rules of effect decomposition¹⁴ (Kline, 2005, p. 129ff), from their joint derivation from a generalized sense of insecurity follows a correlation of r = .40 between concerns about incivility and fear of crime. This association stands facing a direct pathway between the two constructs amounting to r = β = .16. The rather modest direct connecting line is compatible with the classical disorder thesis (Skogan, 1990). The latter assumes that signs of incivility are perceived as cues that signal both the presence of crime and an erosion of informal social control in the neighborhood. Both impressions then increase perceived risk of victimization and emotional worry about crime. Minor fear-enhancing consequences of perceptions of incivility can be observed even after correcting for interdependencies with broader anxieties. In strength, though, they fall conspicuously short of the explanatory power previously ascribed to them. With regard to the differentiation of experiential and expressive fear of crime (Farrall et al., 2009) outlined in the introductory section, one can suppose that the direct linkage between irritation related to incivility and fear of crime depicts the share of the relationship between the two that roots in concrete experienced episodes of worry about incivility and crime.

Discussion

In this article, two theoretical conceptualizations of the connection between anxieties about modernization and fear of crime are examined comparatively: A generalized insecurity perspective that conceives the various fears pervading life in late modernity as immediate materialization of a comprehensive and overarching sense of insecurity is contrasted with an expanded community concern perspective that takes people’s assessment of the “health” of their local neighborhood and community as an indispensable filter mediating the effects of broader anxieties about social change on fear of crime.

First of all, our results support the assumption that the fear of crime is a manifestation of more general fears. Two linear SEMs, one representing a generalized insecurity perspective and one depicting an expanded community concern approach, are both very compatible with survey data from Linz (Austria). For all their differences in detail, both models imply that crime constitutes an outlet for everything that is supposed to go wrong in community and society. Fear of crime forms one of the several interwoven components of a more sweeping sense of insecurity, the origin of which, according to numerous sociological diagnoses of the times, can be found in the political, economic, and social transformations pervading contemporary European societies. As a facet of a complex late-modern sense of insecurity, fear of crime reflects not only the specific fear of falling victim to a crime but also broader perceptions regarding the development of social living conditions in general. This is why fear of crime so often resists attempts at explanation that concentrate on objective risks of crime.

As a result of the basic consistency of both models with the underlying data, the selection of the “better” model, that draws closer to the factual mode of formation of crime-related fear, is difficult. Due to the fact that both models are roughly equivalent in a mathematical sense, similar magnitudes in goodness of fit are to be expected anyway. Hence, the decision which of the two models should be preferred is better based on differences in the complexity of the models and their potential to account for variance in fear of crime, complemented by a close inspection of the substantive findings. These selection criteria advocate a generalized insecurity perspective. The generalized insecurity model requires fewer parameters and enables a better reproduction of the variance in fear of crime than the expanded community concern model. The fact that both economic and social fears exhibit appreciable direct linkages to fear of crime that are not mediated by the perception of local community conditions sheds doubt on the appropriateness of a community concern perspective. In accordance with that, older findings from Austria also portray fear of crime as a materialization of an amorphous sense of insecurity which is projected onto crime (Hirtenlehner, 2006, 2008a; Hirtenlehner & Karazman-Morawetz, 2004).

An additional argument for the superiority of a generalized insecurity perspective can be derived from attempts to bring both models together. Estimating an integrative model once more refers to a predominantly global access to the risks of our times. Expanding the generalized insecurity model with a direct link between worries about incivility and crime makes it clear that the latter does not contribute much to the understanding of the often observed correlations of concerns about incivility and fear of crime when both concepts’ ties to a broader sense of insecurity are controlled for. After correcting for interdependencies with a general syndrome of insecurity, little is left of the well-known fear-enhancing character of perceived disorder.

Of course, while simplicity is desirable, it may be that with a complex phenomenon (such as the fear of crime and its relationship to the other concerns provoked by modern life) simplicity rides roughshod over the nuanced and multifaceted nature of the object of enquiry. While simple solutions are to be encouraged, there may be times when more complex models need to be developed in order to fully represent the subject matter. In addition, while the generalized insecurity model performs slightly better than the expanded community concern model, the latter does suggest a series of causal mechanisms for explaining the fear of crime, whereas the generalized insecurity model has a tendency to disregard causal explanations—tending to “lump” all fears together with little concern about how one may influence another. On the other hand, giving up the assumption of unidirectional cause-and-effect relationships between various fears might also help overcome the limitations of an overly simplistic cognitive psychology. Of course, such matters are empirical ones which cannot be solved in this article.

What are the implications of the present investigation for future research on the mode of formation of fear of crime in Europe? The comparison of two similar and closely related explanatory models whose origins and resonance are located in different cultural spheres provides for deep insight. Gerber, Hirtenlehner, and Jackson (2010, p.152) point the way:

In some countries, public perceptions of crime may operate against the backdrop of broader anxieties concerning long-term social change, but crucially these anxieties translate into fears of crime through mediating perceptions of community and neighbourhood (as found in the UK research of Farrall et al., 2009; Girling et al., 2000; Jackson, 2004). In other countries, fear of crime may be just one facet of this broader set of insecurities (as found in some Austrian and German research, Sessar, 2008; Hirtenlehner, 2006).

This addresses a dependency of the specific mode of formation of fear of crime on the cultural–institutional arrangement of a society. It is possible that fear of crime “relates” more easily to global and national anxieties in the German-speaking and other continental European countries with strong welfare and provision state traditions and solidary-egalitarian value systems than it does in English-speaking countries which embody an individualistic-liberal cultural firmament and a strong orientation on community (Lacey & Zedner, 1998; Schwartz, 2006). While it is the culturally shaped significance of community and neighborhood in English-speaking societies which just might raise the assessment of local conditions to the rank of a major connective link between overall social change and fear of crime, in other countries the dominant etatistic (state-centered) view of the world could open a gateway for a direct translation of social, economic, and existential anxieties into fear of crime. In this respect, a replication of this investigation in Britain could be expected to lead to results which differ from those reported here. The latter appears to be all the more plausible when one takes into consideration the traditionally higher levels of fear of crime and the greater heterogeneity and levels of inequalities among individuals and communities in Britain. Compared to other countries, Austria is characterized by low victimization rates, little fear of crime, and modest inner-city problems (Sessar, Herrmann, Keller, Weinrich, & Breckner, 2004; Van Dijk, van Kesteren, & Smit, 2007). Although the welfare state is also being reconstructed here, Austria still provides a considerable extent of social security and its population has not yet lost confidence in institutional protection against social and economic risks (Hanak, Karazman-Morawetz, & Stangl, 2004). Incivility appears to be actually less common in Austria than in many other countries, but there is some indication that its rare presence is accompanied by a reduced tolerance for heterogeneity and difference and a heightened sensitiveness and irritability concerning signs of disorder (Boers, 2003). Tackling disruptions of inner-city life is traditionally seen as a task of federal and municipal authorities by the citizens and not as a responsibility of the community of residents—a fact that may partially account for the close association of concerns about incivility and social and economic fears (Hanak et al., 2004).

In the future, the sociocultural and political–institutional context will have to be accorded a greater degree of attention, should one want to come to an adequate understanding of the mechanisms and insecurities involved in the formation of fear of crime. We have to be open for the idea that culturally and structurally distinct countries might also give birth to different pathways into fear of crime. Cultural characteristics (e.g., individualism, egalitarism, solidarity, etatism) influence how people make sense of their closer environment and the wider world. They determine what is regarded as risk and who is assigned responsibility to control this risk (Douglas & Wildavsky, 1982). These culturally shaped interpretations may have a considerable impact on the range of anxieties connected with crime, on the linkages between various domains of insecurity, and thereby also on the cognitive processes underlying the formation of fear of crime. Also, a country’s institutional order may play a crucial role. Cross-European comparative research has demonstrated that high levels of economic inequality go together with elevated anxieties about crime (Kristjansson, 2007) and that a great emphasis on social welfare policy or education policy is accompanied by low fear of crime, even when actual rates of victimization are controlled for (Hummelsheim et al., 2011). It has been convincingly argued that a well-developed commercial security industry, a political system shaken by crises of trust or legitimacy and an aggressive media culture pave the way for widespread fear of crime (Farrall et al., 2009; Warr, 2000), whereas corporatist political structures are thought to be associated with lesser fears related to crime (Lappi-Seppälä, 2008). So future research on the roots of fear of crime will be well advised to take into consideration the broader cultural and institutional makeup of a society and their interaction with sentiments of insecurity. We hope to have contributed to the stimulation of a truly comparative research agenda (Nelken, 2010) that takes into account the moderating role of the social, political, and cultural organization of a society when it comes to explaining fear of crime.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Akaike

(1987). Factor analysis and AIC. Psychometrica, 52, 317–332.

Arbuckle

(2007). Amos 16.0 user’s guide. Chicago, IL: SPSS.

Bauman

(1999). In search of politics. Cambridge, England: Polity Press.

Bauman

(2006). Liquid fear. Cambridge, England: Polity Press.

Bauman

(2007). Liquid times. Living an age of uncertainty. Cambridge, England: Polity Press.

Beck

(1992). Risk society: Towards a new modernity. London, England: Sage.

Boers

(1991). Kriminalitätsfurcht. Über den entstehungszusammenhang und die folgen eines sozialen problems. Pfaffenweiler, Germany: Centaurus.

Boers

(2003). Fear of violent crime. In Heitmeyer

Hagan

(Eds.), International handbook of violence research (pp. 1131–1150). Dordrecht, Netherlands: Kluwer.

Brunton-Smith

(2011). Untangling the relationship between fear of crime and perceptions of disorder. Evidence from a longitudinal study of young people in England and Wales. British Journal of Criminology, 51, 885–899.

10.

Byrne

(2001). Structural equation modeling with AMOS. Mahwah, NJ: Lawrence Erlbaum.

11.

Chadee

Virgil

Ditton

(2009). State-trait anxiety and fear of crime. A social psychological perspective. In Lee

Farrall

(Eds.), Fear of crime. Critical voices in an age of anxiety (pp. 168–187). New York, NY: Routledge.

12.

Douglas

Wildavsky

(1982). Risk and culture. An essay of the selection of technological and environmental dangers. Berkeley: University of California Press.

13.

Ewald

(2000). Criminal victimisation and social adaptation in modernity: Fear of crime and risk perception in the new Germany. In Hope

Sparks

(Eds.), Crime, risk and insecurity: Law and order in everyday life and political discourse (pp. 166–199). London, England: Routledge.

14.

Farrall

Jackson

Gray

(2009). Social order and the fear of crime in contemporary times. Oxford, England: Oxford University Press.

15.

Ferraro

(1995). Fear of crime: Interpreting victimization risk. New York, NY: SUNY Press.

16.

Gadd

Jefferson

(2009). Anxiety, defensiveness and the fear of crime. In Lee

Farrall

(Eds.), Fear of crime. Critical voices in an age of anxiety (pp. 125–142). New York, NY: Routledge.

17.

Garofalo

Laub

(1978). The fear of crime: Broadening our perspectives. Victimology, 2, 242–253.

18.

Gerber

Hirtenlehner

Jackson

(2010). Insecurities about crime in Germany, Austria and Switzerland: A review of research findings. European Journal of Criminology, 7, 141–157.

19.

Giddens

(1990). The consequences of modernity. Cambridge, England: Polity Press.

20.

Giddens

(1991). Modernity and self-identity. Cambridge, England: Polity Press.

21.

Girling

Loader

Sparks

(2000). Crime and social change in Middle England. Questions of order in an English town. London, England: Routlege.

22.

Hale

(1996). Fear of crime. A review of the literature. International Review of Victimology, 4, 79–150.

23.

Hanak

Karazman-Morawetz

Stangl

(2004). Insecurities in European cities – Local report Vienna. Vienna, Austria: Institute for the Sociology of Law and Criminology.

24.

Hirtenlehner

(2006). Kriminalitätsfurcht – ausdruck generalisierter ängste und schwindender gewissheiten? Untersuchung zur empirischen bewährung der generalisierungsthese in einer österreichischen kommune. Kölner Zeitschrift für Soziologie und Sozialpsychologie, 58, 307–331.

25.

Hirtenlehner

(2008a). Disorder, social anxieties and fear of crime. Exploring the relationship between incivilities and fear of crime with a special focus on generalized insecurities. In Kury

(Ed.), Fear of crime – Punitivity. New developments in theory and research (pp. 127–158). Bochum, Germany: Brockmeyer.

26.

Hirtenlehner

(2008b). Unwirtlichkeit, unterstützungserwartungen, risikoantizipation und kriminalitätsfurcht. Eine prüfung der disorder-theorie mit österreichischen befragungsdaten. Monatsschrift für Kriminologie und Strafrechtsreform, 91, 112–130.

27.

Hirtenlehner

(2009). Kriminalitätsangst – Klar abgrenzbare furcht vor straftaten oder projektionsfläche sozialer unsicherheitslagen. Journal für Rechtspolitik, 17, 13–22.

28.

Hirtenlehner

(2011). The origins of punitive mentalities in late-modern societies. Testing an expressive explanatory model. In Kury

Shea

(Eds.), Punitivity. International developments. Punitiveness—A global phenomenon? (Vol. 1, pp. 27–51). Bochum, Germany: Brockmeyer.

29.

Hirtenlehner

Karazman-Morawetz

(2004). Hintergründe kriminalitätsbezogener unsicherheitsgefühle. Eine empirisch-statistische analyse am beispiel wiener befragungsdaten. Journal für Strafrecht, 2, 120–123, 161–166.

30.

Hollway

Jefferson

(1997). The risk society in an age of anxiety: Situating fear of crime. British Journal of Sociology, 48, 255–266.

31.

Hough

(2009, June 18). Risk, fear and insecurity: In praise of simplicity. Paper presented at the CRIMPREV Conference, Milton Keynes, UK.

32.

Hummelsheim

Hirtenlehner

Jackson

Oberwittler

(2011). Social insecurities and fear of crime: A cross-national study on the impact of welfare state policies on crime-related anxieties. European Sociological Review, 27, 327–345.

33.

Jackson

(2004). Experience and expression. Social and cultural significance in the fear of crime. British Journal of Criminology, 44, 946–966.

34.

Jackson

(2006). Introducing fear of crime to risk research. Risk Analysis, 26, 253–264.

35.

Jackson

(2009). Bridging the social and the sociological in the fear of crime. In Lee

Farrall

(Eds.), Fear of crime. Critical voices in an age of anxiety (pp. 143–167). New York, NY: Routledge.

36.

Jefferson

Hollway

(2000). The role of anxiety in fear of crime. In Hope

Sparks

(Eds.), Crime, risk and insecurity. Law and order in everyday life and political discourse (pp. 31–49). London, England: Routledge.

37.

Kline

(2005). Principles and practice of structural equation modeling. New York, NY: Guilford Press.

38.

Kristjansson

(2007). On social equality and perceptions of insecurity. A comparison study between two European countries. European Journal of Criminology, 4, 59–86.

39.

Lacey

Zedner

(1998). Community in German criminal justice: a significant absence? Social and Legal Studies, 7, 7–25.

40.

LaGrange

Ferraro

Supancic

(1992). Perceived risk and fear of crime: Role of social and physical incivilities. Journal of Research in Crime and Delinquency, 29, 311–334.

41.

Lappi-Seppälä

(2008). Trust, welfare, and penal culture: Explaining differences in national penal policies. In Tonry

(Ed.), Crime and justice. A review of research (Vol. 37; pp. 313–387). Chicago, IL: University of Chicago Press.

42.

Mohr

(1990). Understanding significance testing. London, England: Sage.

43.

Nelken

(2010). Comparative criminal justice. Making sense of difference. London, England: Sage.

44.

Rengifo

Bolton

(2012). Routine activities and fear of crime: Specifying individual-level mechanisms. European Journal of Criminology, 9, 99–119.

45.

Roberts

Stalans

Indermaur

Hough

(2003). Penal populism and public opinion. Lessons from five countries. Oxford, England: University Press.

46.

Ross

Jang

(2000). Neighborhood disorder, fear, and mistrust: The buffering role of social ties with neighbors. American Journal of Community Psychology, 28, 401–420.

47.

Sampson

Raudenbush

(1999). Systematic social observation of public spaces: A new look at disorder in urban neighborhoods. American Journal of Sociology, 105, 603–651.

48.

Schwartz

(2006). A theory of cultural value orientations: Explication and applications. Comparative Sociology, 5, 137–182.

49.

Schwarz

(1978). Estimating the dimension of a model. Annals of Statistics, 6, 461–464.

50.

Sessar

(1998). Kriminalitätseinstellungen: Von der furcht zur angst? In Schwind

H. D.

Kube

Kühne

(Eds.), Kriminologie an der Schwelle zum 21. Jahrhundert. Festschrift für Hans Joachim Schneider zum 70. Geburtstag am 14. November 1998 (pp. 399–414). Berlin, Germany: de Gruyter.

51.

Sessar

(2003). Unsicherheit als Strukturproblem. In Herrmann

Sessar

Weinrich

(Eds.), Unsicherheit in der moderne am beispiel der grossstadt. Jahrbuch ’02 für rechts- und kriminalsoziologie (pp. 252–261). Baden-Baden, Germany: Nomos.

52.

Sessar

(2008). Fear of crime or fear of risk? Some considerations resulting from fear of crime studies and their political implications. In Kury

(Ed.), Fear of crime – Punitivity. new developments in theory and research (pp. 25–32). Bochum, Germany: Brockmeyer.

53.

Sessar

(2010). Kriminalitätseinstellungen und sozialer wandel. Gleichzeitig eine auseinandersetzung mit forschungen zur verbrechensfurcht und punitivität. Monatsschrift für Kriminologie und Strafrechtsreform, 93, 361–381.

54.

Sessar

Herrmann

Keller

Weinrich

Breckner

(2004). Insecurities in European cities. Crime-related fear within the context of new anxieties and community-based crime prevention. Hamburg, Germany: Department of Criminology of the University of Hamburg.

55.

Skogan

(1990). Disorder and decline. Crime and the spiral of decay in American neighborhoods. Berkeley: University of California Press.

56.

Smith

Pain

(2009). Critical geopolitics and everyday fears. In Lee

Farrall

(Eds.), Fear of crime. Critical voices in an age of anxiety (pp. 45–58). New York, NY: Routledge.

57.

Taylor

Evans

Fraser

(1996). A tale of two cities. Global change, local feeling and everyday life in the north of England. London, England: Routledge.

58.

Taylor

Jamieson

(1998). Fear of crime and fear of falling. English anxieties approaching the millennium. Archives Européennes de Sociologie, 39, 149–175.

59.

Tulloch

Lupton

Blood

Tulloch

Jennett

Enders

(1998). Fear of crime. Canberra, Australia: CCCRS.

60.

Van Dijk

van Kesteren

Smit

(2007). Criminal victimization in international perspective. Key findings from the 2004–2005 ICVS and EUICS. The Hague, Netherlands: WODC.

61.

Vilsmeier

Taschler-Polacek

(1991). Viktimisierungsangst, der “ruf nach recht und ordnung” und allgemeine ängstlichkeit – eine umfrage unter älteren frauen. Monatsschrift für Kriminologie und Strafrechtsreform, 74, 174–181.

62.

Vitelli

Endler

(1993). Psychological determinants of fear of crime: A comparison of general and situational prediction models. Personality and Individual Differences, 14, 77–85.

63.

Walklate

Mythen

(2008). How scared are we? British Journal of Criminology, 48, 209–225.

64.

Warr

(2000). Fear of crime in the United States. Avenues for research and policy. Criminal Justice, 4, 451–489.

65.

Fiedler

Flaming

(2005). Discovering the impact of community policing: The broken windows thesis, collective efficacy, and citizens judgment. Journal of Research in Crime and Delinquency, 42, 147–186.

66.

Zarafonitou

(2009). Criminal victimization in Greece and the fear of crime: A ‘paradox‘ for interpretation. International Review of Victimology, 16, 277–300.